---
title: "Researchers Test Stargate Firewall Against 11 CubeSat Network Attacks"
id: 227311
type: "post"
slug: "stargate-firewall-cubesat-network-attacks"
author: "NSE"
published: "2026-11-21T00:49:31+00:00"
modified: "2026-11-20T13:52:48+00:00"
canonical: "https://newspaceeconomy.ca/2026/11/20/stargate-firewall-cubesat-network-attacks/"
markdown_url: "https://newspaceeconomy.ca/2026/11/20/stargate-firewall-cubesat-network-attacks.md"
description: "Researchers tested a firewall that checks CubeSat commands and contact windows, blocking 11 attack scenarios in a virtual mission network."
language: "en-US"
category:
  - "Current News"
  - "Space Economy"
post_tag:
  - "Cubesats"
  - "Cybersecurity"
  - "Satellite Communications"
  - "Satellite Networks"
  - "Space Security"
---

# Researchers Test Stargate Firewall Against 11 CubeSat Network Attacks

Researchers at New Mexico State University and Universidad de Santiago de Chile have developed a [CubeSat](https://newspaceeconomy.ca/2025/08/25/cubesat-design-specification-csd-an-overview/) firewall that blocked all 11 attack scenarios in a virtual spacecraft network. Their Stargate system checks individual spacecraft commands and authorized contact periods, adding 1.42 milliseconds to median round-trip communication time in the reported experiments.

Owen Ramsey, Carlos Gonzalez-Cortes, and Efrén López-Morales described the results in a [paper posted November 13, 2026](https://arxiv.org/abs/2611.18874). The manuscript is listed for forthcoming publication in the proceedings of the first IEEE International Conference on Resilience and Integrated Security for Space and Critical Systems. The team tested the firewall with real open-source flight software running inside a virtual mission environment.

## Checking Commands Before They Reach a Spacecraft

Stargate operates at the boundary between the ground network and spacecraft communications. It examines messages using the CubeSat Space Protocol, or CSP, and decides whether to forward or reject them according to rules supplied by the mission. Those rules can distinguish the sender, the requested service, the particular command, and whether the satellite is within an authorized communication window.

The [CSP project documentation](https://libcsp.github.io/libcsp/) describes a compact communications system designed for computers with limited processing power and memory. It lets spacecraft subsystems exchange messages over a shared network. A message contains addressing information that identifies its destination and the service being requested, allowing equipment to communicate without routing every exchange through a single master controller.

That compact design makes the contents of a command relevant to the firewall decision. Several different spacecraft commands can use the same service port. Allowing traffic to that port alone can permit both a routine operation and a more privileged instruction. Stargate reads the command inside the message so that the two can receive different treatment.

The researchers demonstrated this distinction by sending an allowed command and a restricted command with the same source, destination, and service port. The firewall forwarded the allowed instruction and rejected the restricted one. The decision depended on the requested operation rather than the network address alone.

## Four Types of Rules

The [study organizes the tested attacks](https://arxiv.org/html/2611.18874v1) into four categories. Network identity rules control which nodes can communicate, which services they can reach, and which message priorities or traffic rates are allowed. Management rules restrict administrative operations. Command rules examine application instructions, and mission rules add operational conditions such as an authorized contact period.

In the network tests, an authorized ground-station node could communicate with the spacecraft, but traffic from an unauthorized node was rejected. The same group of experiments examined access to restricted services, misuse of packet priority, and traffic that exceeded the configured command rate. These checks were applied before the messages entered the spacecraft-side network.

The management tests covered five operations that could read memory, change memory, alter routing information, adjust the system clock, or restart a target node. The researchers first established each operation’s effect without the relevant policy. Repeating the tests with the policy enabled blocked all five unauthorized operations.

The contact-window test used the same otherwise permitted command at different times. Stargate accepted it during an authorized communication period and rejected it outside that period. The prototype calculated satellite visibility with the Skyfield software library and publicly available orbital data from CelesTrak. This supplied timing information that the packet itself did not contain.

The team implemented Stargate in C using the libcsp library, with mission rules stored in a configuration file. Its processing sequence separates reading packets, interpreting their contents, evaluating policies, and enforcing the resulting decision. That separation allows mission-specific command interpretation and rules to be changed without replacing the entire firewall.

## Testing With Flight Software

The virtual test environment represented an authorized ground station, an unauthorized ground node, the firewall, a satellite computer, and an additional onboard payload. The satellite node ran open-source flight software with CSP support. The firewall sat between the ground-side and spacecraft-side networks so that traffic directed toward the spacecraft passed through its checks.

Experiments ran in an Ubuntu virtual machine with eight virtual processors and 16 gigabytes of memory. For each attack, the researchers compared operation with the relevant policy disabled and enabled. All 11 selected scenarios were blocked when the corresponding protections were active. The implementation and evaluation used CSP version 1.

The timing measurements compared CSP ping exchanges with and without Stargate. Median round-trip time rose from 10.83 milliseconds to 12.25 milliseconds. Mean round-trip time increased from 51.76 milliseconds to 63.36 milliseconds. Those measurements describe end-to-end communication in the virtual environment, including its software and container overhead.

The distinction between network access and spacecraft-command authorization connects the study to earlier [satellite cybersecurity assessments](https://newspaceeconomy.ca/2024/02/26/report-leo-satcom-cybersecurity-assessment-eu-2024/) covering both ground and space systems. Stargate applies that distinction to CSP traffic at the ground-space boundary, where an instruction can be checked before it reaches the target flight software.

The authors have also identified a public repository for the firewall source code and experimental results in their paper.

![Infographic describing 11 blocked CubeSat attack scenarios in virtual tests and a 1.42-millisecond increase in median round-trip time.](https://i0.wp.com/newspaceeconomy.ca/wp-content/uploads/2026/11/stargate-cubesat-firewall.png?resize=696%2C464&ssl=1)